Re: C-x 2 and C-x 3

[David De La Harpe Golden]

On 26/10/11 15:06, Stefan Monnier wrote:
> This said, I agree with Michael that it'd be good if the name made it
> clear which window is the new one

On 26/10/11 15:58, Drew Adams wrote:
> That said, it seems that which window is the new one or the selected one has
> become (more?)important now.    If so,

Hit the nail on the head there. I'm not clear myself why it's now more 
important (and if so, as you say). It kind of seems like an 
implementation detail that the original window survives at all - in a 
different implementation that'd be functionally quite similar to the 
user, post-call _both_ windows could be new "copy" windows in 
object-identity terms, with the space that the original window took up 
now split between them and the selected window now the left or top one 
(Yes, the return value is defined to be "the" new window currently, but 
again, something of an implementation detail...)

If one wanted to hide (at least from the function name) which windows 
are original (if any) and which copies, well,

split-window-side-by-side-with-selected-on-left
split-window-one-above-the-other-with-selected-on-top

or could revert to -horizontally/-vertically, since the -select suffix
also happens to effectively disambiguate:

split-window-horizontally-select-left
split-window-vertically-select-top

(and make
split-window-horizontally-select-right
split-window-vertically-select-bottom
if anyone feels the need or symmetry compulsion)